JOURNAL OF BACTERIOLOGY, JUlY 1988, p. 3072-3079 Vol. 170, No. 70021-9193/88/073072-08$02.00/0Copyright C 1988, American Society for Microbiology

Autotrophic Acetyl Coenzyme A Biosynthesis inMethanococcus maripaludis

JERSONG SHIEHt AND WILLIAM B. WHITMAN*

Department of Microbiology, University of Georgia, Athens, Georgia 30602

Received 1 February 1988/Accepted 12 April 1988

To detect autotrophic CO2 assimilation in cell extracts of Methanococcus maripaludis, lactate dehydrogenaseand NADH were added to convert pyruvate formed from autotrophically synthesized acetyl coenzyme A tolactate. The lactate produced was determined spectrophotometrically. When CO2 fixation was pulled in thedirection of lactate synthesis, CO2 reduction to methane was inhibited. Bromoethanesulfonate (BES), a potentinhibitor of methanogenesis, enhanced lactate synthesis, and methyl coenzyme M inhibited it in the absence ofBES. Lactate synthesis was dependent on CO2 and 112, but H2 + C02-independent synthesis was also observed.In cell extracts, the rate of lactate synthesis was about 1.2 nmol min-' mg of protein-'. When BES was added,the rate of lactate synthesis increased to 2.3 nmol min-' mg of protein-'. Because acetyl coenzyme A did notstimulate lactate synthesis, pyruvate synthase may have been the limiting activity in these assays. Radiolabelfrom '4CO2 was incorporated into lactate. The percentages of radiolabel in the C-1, C-2, and C-3 positions oflactate were 73, 33, and 11%, respectively. Both carbon monoxide and formaldehyde stimulated lactatesynthesis. 14CH20 was specifically incorporated into the C-3 of lactate, and 14CO was incorporated into the C-1and C-2 positions. Low concentrations of cyanide also inhibited autotrophic growth, CO dehydrogenaseactivity, and autotrophic lactate synthesis. These observations are in agreement with the acetogenic pathwayof autotrophic CO2 assimilation.

Most species of methanogenic bacteria can oxidize hydro-gen and reduce carbon dioxide to methane to obtain energyfor growth (26, 47). In addition, about one-half of the speciesof methanogens which have been described are able to growautotrophically. Nevertheless, the pathway of CO2 fixationduring autotrophic growth in methanogens has not been fullyestablished, and the carbon metabolism of only a few specieshas been characterized. In Methanobacterium thermoauto-trophicum and Methanosarcina barkeri, a novel pathway ofacetyl coenzyme A (acetyl-CoA) synthesis from two mole-cules of CO2 has been proposed (16, 21, 29, 42-44). Thispathway closely resembles the total synthesis of acetatedescribed previously in Clostridium thermoaceticum andother anaerobic eubacteria (8, 13, 22, 24, 25, 32, 33, 50).The acetyl-CoA pathway in C. thermoaceticum involves

total synthesis of acetate from two molecules of CO2 viadifferent routes of reduction (32, 50). One molecule of CO2 isreduced to the methyl level via the folate pathway, and theother molecule of CO2 is reduced to a bound CO residue.They are then condensed to an activated acetyl residuewhich is thiolized with HS-CoA. The final step of acetyl-CoA biosynthesis is catalyzed by CO dehydrogenase (36).

In Methanobacterium thermoautotrophicum, acetyl-CoAis an early product of CO2 assimilation (15, 17, 38). Labelingstudies have shown that the biosynthesis of acetyl-CoA inMethanobacterium thermoautotrophicum resembles the to-tal synthesis of acetyl-CoA from 2CO2 by acetogens. Themethyl group of acetate comes from methyltetrahydrome-thanopterin (methyl-THMPT), which is a folate analog andan intermediate in the reduction of CO2 to methane (11, 28,30, 45). The carboxy group of acetyl-CoA is derived fromanother molecule of C02, which is reduced to a bound CO byCO dehydrogenase, or it can come directly from carbon

* Corresponding author.t Present address: Michigan Biotechnology Institute, Lansing,

MI 48909.

monoxide gas (46). Thus, autotrophic CO2 assimilation andCO2 reduction to methane have intermediates in common(21).

In this report, we describe a spectrophotometric assay forthe acetogenic pathway in cell extracts of Methanococcusmaripaludis. Using this assay, we also demonstrated thepresence of this pathway for the first time in the Methano-coccales, one of three major orders of methane-producingbacteria (26, 51).

MATERIALS AND METHODSBacteria and growth conditions. Methanococcus maripa-

ludis JJ was grown anaerobically on H2-C02 (80:20 [vol/vol])gas at 37°C in mineral medium as described previously (27,48). A typical fermentor was prepared as follows. Ten litersof mineral medium was autoclaved under N2 gas for 40 min,and 20 ml of a sterile solution containing 10 g of 2-mercap-toethanesulfonate and 5 ml of a sterile solution of 20% (wt/vol) Na2S 9H20 was then added. The fermentor was inoc-ulated with 1 liter of culture. Sterile 20% Na2S 9H20 (5 ml)was added to the fermentor twice a day during growth. Thepressure of H2 + CO2 gas was maintained at 100 kPa, and theflow rate was 100 to 250 ml/min. Cells were harvested in theearly stationary phase with a Sharples continuous-flow cen-trifuge and stored under H2 gas at -20°C.

Preparation of extracts. Cell pellets were thawed in ananaerobic chamber (Coy Laboratory, Ann Arbor, Mich.).Cells were lysed by addition of 25 mM potassium-PIPES[piperazine-N,N'-bis(2-ethanesulfonic acid)] buffer, pH 6.8,containing 1 mM dithiothreitol and 1 mM cysteine, whichwas saturated with H2 gas. Pancreatic DNase, 1 mg/20 g (wetweight) of cells, was added. The cell extract became homo-geneous after 30 min of incubation in the chamber at roomtemperature, and it was then centrifuged at 30,000 x g for 30min at 4°C in a polypropylene centrifugation bottle whichhad been equilibrated in an anaerobic chamber for at least 1day. The supernatant was collected in an anaerobic chamber

3072

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

0METHANOCOCCUS AUTOTROPHY 3073

and was either used immediately or stored under H2 gas at-20°C. Under these conditions, activity was retained for 6months. The protein contents of extracts were determinedby the method of Lowry et al. (34) after digestion in 0.1 MNaOH for 30 min at 90°C. Similar results were obtained bythe dye-binding method (3).

In vitro lactate synthesis. Assays were performed in 2.6-mlserum vials with butyl rubber stoppers and aluminum seals(West Co., Inc., Phoenixville, Pa.). Before the assays,extracts were preincubated under H2 gas to remove endog-enous substrates. Typically, 4 ml of extract was incubatedfor 1 h at 37°C in a 160-ml serum bottle under an atmosphereof HJ-2 at 200 kPa. Lactate dehydrogenase (LDH) and othercomponents of the assay were prepared in 50 mM Tricine[N-Tris(hydroxymethyl)methyl glycine] buffer (pH 7.5;Sigma Chemical Co., St. Louis, Mo.). Except for LDH,oxygen-stable components were dispensed in small volumes(usually 10 ,ul) into the assay vials outside the anaerobicchamber. During passage through the air lock into thechamber, the solutions were maintained in vacuo for 5 minduring one of the three normal cycles. The vials were thensealed inside the chamber. Stoppers for the vials were takeninto the chamber at least 2 days before use. LDH wasprepared separately in a total volume of 0.4 ml in a serumvial. The vial was stoppered, and the stopper was puncturedwith a 22-gauge needle. In this way, the gas in the vial wasexchanged with N2 gas during passage through the air lock tothe anaerobic chamber. Upon removal from the anaerobicchamber, the assay vials were placed in an ice bath andflushed with H2 gas for 5 min at a flow rate of 10 ml/min.Flushing with H2 continued during addition of extracts andLDH. Cell extract (0.2 ml containing 8 mg of protein) wasdispensed into the assay vials with a 1-ml Glas-pak dispos-able syringe (Becton Dickinson Vacutainer Systems, Ru-therford, N.J.). After addition of the extracts, LDH (10 RIcontaining 25 U) and bromoethanesulfonate (BES; 10 ,ulcontaining 20 nmol) were added to the vials with a 50 ,u1syringe (Hamilton Co., Reno, Nev.). Upon addition of thelast component, the vials were flushed with H2-C02 (80:20[vol/vol]) for 5 min. The assays were initiated by transfer ofthe vials to a 37°C water bath. After 40 min, the vials weretransferred to an ice bath and a 20-,ul gas sample wasremoved from each vial with a 100-,l gas-tight syringe(Precision Sampling Corp., Baton Rouge, La.) and analyzedfor methane by gas chromatography (49). The extracts werethen acidified by addition of 0.2 ml of 60% HCl04, trans-ferred to 1.5-ml centrifuge tubes, and centrifuged for 4 min at10,000 x g at room temperature. After neutralization of 0.2ml of the sUpernatant with 0.1 ml of 1.5 M K2C03, lactatewas measured spectrophotometrically with LDH and NAD+(20).

Pyruvate synthase was measured by a similar procedure.The only difference was that 20 pumol of acetyl phosphateand 5 U of phosphotransacetylase were included in theassays under H2-CO2.

All assays were performed in duplicate, and the resultsshown are the averages. In addition, results were repro-duced at least once. In most cases, extracts from more thanone batch of cells were tested. When cyanide was used, thepH of the assay was increased by 0.4 U with 20 mM KCN.

Incorporation of radiolabel into lactate. In radiolabelingexperiments, 14C02, 14CO, and 14CH20 were added to theassays after removal of the gassing needles. Thle specificactivity of the 14CO2 was determined from a sample of thegas phase. The quantity of CO2 was determined by gaschromatography with a thermal conductivity detector (Va-

rian 3760; Sugarland, Tex.) under the same conditions usedto analyze CH4. Radioactivity was measured by liquidscintillation counting after trapping of the radiolabel inC02-trapping cocktail. 14CO was prepared from radiolabelformate (14). The specific activity of 14CO was calculatedfrom the specific activity of the labeled formate, assumingcomplete conversion. The specific activities of 14CO2, 14CO,and 14CH2O used in these experiments were 6.64 x 106, 1.36x 106, and 6.83 x 106 dpm/lpmol, respectively. At theconclusion of the radiolabeling experiment, extracts wereacidified as described above. After centrifugation, a portionof the supernatant was assayed for lactate. Unlabeled lactate(1 mmol) was added to the remaining supernatant as acarrier.

Purification and degradation of lactate. Lactate was puri-fied by chromatography on Celite (Johns-Manville ProductsCorp., Manville, N.J.) and Dowex 5OW, H+ (1, 18). Theradiochemical purity of lactate was confirmed by thin-layerchromatography on silica gel plates with chloroform-i-butanol (80:20) and fluorography (2). On KB silica gel plates(Whatmnan Inc., Clifton, N.J.), the Rf of lactate was 0.14.Before the degradation, an additional 1 mmol of unlabeledlactate was added. C-1 degradation of lactate was performedby dichromate oxidation (1). Acetate, which is the product,was purified from the lysate by steam distillation, taken todryness, suspended in about 1 ml of distilled water, anddegraded further by the Schmidt method (1). Total degrada-tion of lactate to 3CO2 was performed by the Van Slyke-Folch oxidation procedure (39). The 14CO2 released fromeach degradation step was absorbed in a 2 N NaOH solutionwhose weight had been carefully predetermined. Theamount of CO2 trapped in the NaOH solution was thencalculated from the increase in weight. The amount ofradiolabel trapped was determined by liquid scintillationcounting.Formation of '4Co2 from radiolabeled substrates. The rate

of formation of 14CO2 from 14CH20 and 14CO was deter-mined under the conditions of the standard assay, exceptthat the volume was increased. In a 10-ml serum bottle, 0.5ml of extract (18 mg of protein) was incubated with 6 p.mol ofNADH and 60 U of LDH under an atmosphere of H2-CO2(80:20 [vol/vol]). Radiolabeled 14CH20 (2.5 ,umol containing5.6 x 106 dpm) or 14Co (0.4 mmol containing 5.1 x 106 dpmwas then added. After 40 min at 37°C, the gas inside thebottle was flushed out by a stream of N2 gas into CO2-trapping cocktail, and the radioactivity was determined byliquid scintillation counting.Enzymatic assays. CO dehydrogenase was assayed in 1 ml

of 50 mM phosphate buffer, pH 7.5, which contained 20 mMmethyl viologen, at 25°C. The cuvettes were stoppered withred serum rubber stoppers and made anaerobic by repeatedvacuuming and flushing with N2 gas. The cuvettes were thenflushed with 100% CO gas for 30 s and pressurized to 100kPa. The CO gas was made oxygen free by beitng flushedthrough a solution of 25 mM methyl viologen, which wasreduced with about 10 mM sodium dithionite. Enough so-dium dithionite was then added to the cuvettes to turn thereaction mixture slightly blue. The assay was initiated byinjection of the cell extract. The reduction of methyl violo-gen was monitored at 603 nm with a Gilford spectrophotom-eter. The extinction coefficient of methyl viologen was 11.3x 103 cm-1 M-1. The activity was linear with protein from0.2 to 0.8 mg. Pyruvate oxidoreductase was assayed in 100mM potassium-Tricine buffer, pH 8.6, which contained 20mM methyl viologen. The cuvettes were prepared in themanner used for the CO dehydrogenase assay except that

VOL. 170, 1988

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3074 SHIEH AND WHITMAN

TABLE 1. LDH trap for autotrophic acetyl-CoA synthesis

a Lactate CH4,formedAddition(s) (amt) AA formed (nmol) (nmol)

Noneb 0.052 0 2,800NADH (2.5 ,umol) 0.031 0 3,700LDHI (25 U) 0.060 11 3,100NADH + LDH 0.330 388 1,400

a Change in A340 observed in the spectrophotometric assay for lactate.b With no additions, the assay included 8 mg of protein under an atmo-

sphere of H2 + CO2.

the gas phase was N2. After addition of cell extract, thereaction was initiated by addition of sodium pyruvate (finalconcentration, 10 mM) and CoA (final concentration, 0.1mM). Reduction of methyl viologen was again monitored at603 nm. The reaction was both pyruvate- and CoA depen-dent, and it was linear with protein from 0.2 to 0.8 mg.

Materials. All chemicals were reagent grade or better. H2and H2-CO2 (80:20 [vol/vol]) were purchased from Selox Co.(Gainesville, Ga.). Carbon monoxide was obtained fromFisher Scientific (Springfield, N.J.). Rabbit muscle lactatedehydrogenase (type XI), bovine pancreatic DNase (typeIV), NAD+, NADH, acetyl-CoA, acetyl phosphate, ATP,ADP, AMP, pyruvate, lactate, PIPES, and Tricine bufferswere obtained from Sigma. BES was obtained from AldrichChemical Co., Inc. (Milwaukee, Wis.). NaH14CO3, 14CH20(30 mnCi/mmol), and [14C]formate (40 mCi/mmol) were ob-tained from ICN Radiochemicals (Irvine, Calif.). Methylcoenzyme M (methyl-CoM) was prepared as described pre-viously (49). 14CO2 gas was prepared from NaH14CO3. Asolution containing 8.4 mg of NaH'4C03 (5 mCi/mmol) in 5ml of sterile water was injected with a syringe into a 10-mlserum bottle containing N2 gas at 50 kPa. One milliliter of 6M HCI was then added with a syringe, and the solutionswere gently mixed. The gas phase was transferred with ashort length of tubing to a second serum bottle which hadpreviously been evacuated. The first bottle was then filledwith a saturated solution of NaCl to effect complete transferof the gas phase to the second bottle. After disconnection ofthe tubing between the two bottles, 2 ml of 2.5% (wt/vol)Na2S - 9H2O was added to the bottle containing the 14CO2.The bottle was then incubated on a rotary shaker for 24 h toeliminate contamination by 02. C02-trapping cocktail con-tained 0.5 liter of toluene, 0.4 liter of methanol, 0.1 liter ofethanolamine, and 2 g of 2,5-bis[5'-tert-butylbenzoxazolyl-(2')]thiophene.

RESULTS

Spectrophotometric assay for acetyl-CoA synthesis. In cellextracts, acetyl-CoA synthesis was barely detectable, pre-sumably because acetyl-CoA was rapidly metabolized toother compounds. Therefore, an assay was devised to trappyruvate formed from autotrophically synthesized acetyl-CoA and pyruvate synthase as lactate by addition of LDHand NADH. Extracts of Methanococcus maripaludis syn-thesized lactate when supplemented with LDH and NADHunder an H2-CO2 atmosphere (Table 1). A small change inabsorbance was also observed during the spectrophotomet-ric determination of lactate in extracts incubated withoutthese components. This change was also observed whenLDH was omitted from the subsequent spectrophotometricassay for lactate. Therefore, interference by chromophoresin the extract limited the sensitivity of the assay. Althoughthese chromophores could be removed by dialysis, lactate

TABLE 2. Effect of gas atmosphere during the assay andpreincubation on lactate synthesis and methanogenesis

Atmosphere Atmosphere of Lactate CH4of assay preincubationa synthesis formation(nMol)b (nmol)c

H2-C02 H2 239 2,580N2-C02 H2 167 310H2 H2 176 390N2 H2 149 290H2-C02 N2 145 1,450N2-C02 N2 151 60H2 N2 150 230

a Extracts were preincubated under H2 or N2 for 60 min.b The assay included 2.5 mg of protein, 2.5 ,umol of NADH, 25 U of LDH,

and 0.1 mM BES. The assay lasted 40 min.'CH4 formation without LDH, NADH, and BES.

synthesis activity was also greatly reduced (data not shown).Under the standard assay conditions, lactate synthesis waslinear with the amount of extract added, and exogenouspyruvate was recovered almost quantitatively (data notshown). The temperature optimum for lactate synthesis was37°C or close to the growth temperature of Methanococcusmaripaludis. The loss in activity above 37°C was not due toinactivation of the LDH, which retained activity up to 55°C.These results established the conditions for lactate synthesisin extracts of Methanococcus maripaludis.

Lactate synthesis was dependent on the gas atmosphere ofthe assay and the preincubation. Lactate synthesis in crudeextracts was sensitive to 02- Subsequent to exposure of theextracts to air for 5 min, lactate synthesis was not observed(data not shown). This result was consistent with the ex-treme oxygen sensitivity of the enzymes of acetyl-CoA andpyruvate synthesis (52). Extracts were normally preincu-bated for 1 h before the assay to reduce the endogenoussubstrates for lactate synthesis. After preincubations underN2, lactate synthesis and methane production when assayedunder H2-C02 were reduced to about 60% of the activityafter preincubation with H2 (Table 2). Presumably, prein-cubation under H2 was necessary to conserve the reducingability of the extracts. The amounts of lactate formed whenthe assays were performed under N2-C02 or H2 was reducedto about 70% of the activity under H2-C02 (Table 2). Thus,lactate synthesis was partially dependent on both H2 andCO2. Because lactate synthesis was observed in the absenceof C02, endogenous substrates were probably a source ofsome of the lactate.

In extracts of Methanobacterium thermoautotrophicum,methyl-CoM stimulates acetyl-CoA synthesis from C02,presumably because it stimulates CO2 reduction to methyl-THMPT and CH4 (19, 30, 37, 44). Likewise, the inhibitor ofthe methyl-CoM reductase system, BES, inhibits acetyl-CoA synthesis because it prevents reduction of CO2 (44). Inextracts of Methanococcus maripaludis, neither NADH norLDH alone inhibited methanogenesis from H2-C02, butwhen they were combined methanogenesis was inhibited by50% (Table 1). This observation supported the hypothesisthat lactate synthesis competes for an intermediate of me-thanogenesis. However, in extracts of Methanococcus ma-ripaludis, methyl-CoM inhibited lactate synthesis, and BESstimulated lactate synthesis (Table 3). Therefore, the rate oflactate synthesis was probably not limited by the availabilityof C-1 units from methanogenesis. Alternatively, both lac-tate synthesis and methanogenesis required reducing equiv-alents from H2. Even though H2 was present in great excess,electron transfer may have been disrupted during prepara-

J. BACTERIOL.

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

METHANOCOCCUS AUTOTROPHY 3075

TABLE 3. Effects of methyl-CoM and BES onlactate synthesis

Addition(s) (amt) Lactate CH4 formedformed (nmol) (nmol)

Nonea 420 1,300BES (0.1 mM) 600 200Methyl-CoM (1 ,umol) 330 2,100BES + methyl-CoM 700 200

a With no additions, the assay included 8 mg of protein, 2.5 ,umol ofNADH,25 U of LDH, and H2-CO2 gas.

tion of the extracts. Therefore, the availability of a commonreductant may have been limiting.

Lactate synthesis was linear with time for at least 1 h (Fig.1). In the presence or absence of BES, the specific activity oflactate synthesis was 2.3 or 1.2 nmol min-' mg of protein-',respectively. The C02-dependent lactate synthesis was alsoproportional to the CO2 concentration in the assay (Fig. 2).The concentration of CO2 for 50% activity was about 2%.When acetyl-CoA was generated in the assay, the concen-tration ofCO2 required for 50% activity of pyruvate synthasewas 1%. At CO2 concentrations higher than 5%, the maximalspecific rate of C02-dependent pyruvate synthesis in thisassay was about 1.2 nmol min-1 mg of protein-1. This shows

2

0

E

I

0.4

E

0)-

0.20

20 40

Time (min.)FIG. 1. Rate of lactate and methane synthesis from H2-CO2. The

assays were performed under standard conditions in the presence(0) or absence of (0) of 0.1 mM BES.

1600

w

-J

[C02] (%v/V)FIG. 2. Effects of CO2 and acetyl-CoA on lactate synthesis.

Extracts, 4 mg of protein, were assayed under standard conditionsas described in the text. Symbols: 0, without acetyl-CoA; A, withacetyl-CoA-generating system. C02-independent lactate synthesiswas subtracted.

that the maximal specific rate of C02-dependent acetyl-CoAformation in this assay is .1.2 nmol min1 mg of protein-'.

Radiolabel from 04C02 was incorporated into lactate(Table 4). Most of the radiolabel was incorporated into theC-1 position. This result would be expected if pyruvate wasformed from acetyl-CoA and CO2 by pyruvate synthase (40).Somewhat less radiolabel, 45 and 15% of the amount in theC-1 position, was found in the C-2 and C-3 positions oflactate, respectively (Table 4). The uneven labeling of theC-2 and C-3 positions could have resulted from acetyl-CoAsynthesis from endogenous C-1 units and CO2 or an ex-change reaction of CO2 and the C-1 of acetyl-CoA (C-2 oflactate) (36).Pathway of CO2 assimilation. In other methanogens,

THMPT and CO are C-1 donors for autotrophic acetyl-CoAbiosynthesis (30, 31). CH20 condenses with THMPT to formmethylene-THMPT in extracts, and it is an intermediate inmethanogenesis or incorporated into the methyl group ofacetyl-CoA (12, 31). CO is incorporated into the carboxygroup. In extracts of Methanococcus maripaludis, CH2Owas converted quantitatively to methane (data not shown).Thus, CH20 appeared to have the same mode of action asfound in other methanogens (12). In the presence of BESadded to inhibit methanogenesis, CH20, and CO stimulated

TABLE 4. Degradation of radiolabeled lactate formedfrom 14CO2

Carbon atom(s) Specific radioactivity of % of radiolabelof lactate CO2 (103 dpm/mmol) in lactatea

All 24.7C-1 54.0 73C-2 24.1 33C-3 8.0 11

a The total (100%) radiolabel in lactate was equal to three times the specificactivity of the CO2 obtained by the total degradation of lactate. Thus, theenrichment of each carbon was obtained from the relative specific radioactiv-ity of the CO2 released during degradation.

VOL. 170, 1988

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3076 SHIEH AND WHITMAN

TABLE 5. Stimulation of lactate synthesis by CH2O and CO

Addition(s) Lacate CH4 formed(amt or pressure) formed (nmol) (nmol)

Nonea 510 100CH20 (1 ,umol) 520 90CO (100 kPa) 620 60CH2O + CO 730 90

a With no additions, the assay included 8 mg of protein, 2.5 iLmol ofNADH,25 U of LDH, 0.1 mM BES, and H2-C02 gas.

lactate synthesis (Table 5). Although the stimulation wasonly about 40%, it was very reproducible and was obtainedwith extracts from at least five batches of cells. 14CH20 and14CO specifically radiolabeled the C-3 and C-2 carbons oflactate, respectively (Table 6). For 14CH20, 71% of the totalradiolabel was incorporated into the C-3 carbon of lactate,which corresponded to the methyl carbon of acetyl-CoA. Todetermine whether incorporation of radiolabel into carbonsC-1 and C-2 was due to oxidation of '4CH20 to 04C02, therate of this reaction was measured. Under the conditions ofthe lactate synthesis assay, extracts converted 14CH20 to14CO2 at a rate of 0.87 nmol min-1 mg of protein-1. Thus,about one-third of the initial 14CH20 was converted to 14C02by the end of the assay. For 14CO, only small amounts ofradiolabel were incorporated. However, the C-2 of lactate,which corresponded to the carboxy of acetyl-CoA, wasspecifically enriched (Table 6). Most of the remaining radio-label was found in the C-1 of lactate, which was probablyderived from CO2 (Table 4). Like CH2O, 14CO was con-verted to 14CO2 by these extracts. Under the conditions ofthe lactate synthesis assay, 14CO was converted to 14CO2 ata rate of 0.2 nmol min-1 mg of protein-'.

In acetogenic bacteria, the immediate methyl donor foracetyl-CoA synthesis is a corrinoid protein (7, 23). In Me-thanobacterium thermoautotrophicum, methyl iodide is amethyl donor, presumably because it chemically methylatesa corrinoid intermediate (21). In extracts of Methanococcusmaripaludis, methyl iodide also stimulated lactate synthesisin the presence of CO (Table 7). The stimulation wasreproducible, and it was observed in extracts from twobatches of cells. This observation suggested that a corrinoidprotein is an intermediate in acetyl-CoA synthesis in Metha-nococcus maripaludis as well.The pathway of autotrophic CO2 assimilation in Methano-

coccus maripaludis appeared to resemble closely that foundin other methanogens and the acetogenic eubacteria. Thisconclusion was further confirmed by detection of high spe-cific activities of enzymes usually associated with this path-way (Table 8). The CO dehydrogenase activity, assayed in

TABLE 6. Degradation of radiolabeled lactate formed from14CH20 and 14CO

Specific radioactivity (103 dpm/mmol) ofCarbon atom(s) CO2 (% of radiolabel) with:

of lactate14CH20 14CO

All 9.4 0.50C-i 4.0 (14) 0.71 (47)C-2 2.3 (8) 0.76 (51)C-3 20.1 (71) 0.19 (13)

a The total (100%) radiolabel in lactate was equal to three times the specificactivity of the CO2 obtained from the total degradation of lactate (see Table 4footnote).

TABLE 7. Stimulation of lactate synthesis by methyl iodideand CO

Addition(s) Lactate CH4 formed(amt or pressure) formed (nmol) (nmol)

None" 480 180CH31 (5 mM) 540 10CO (100 kPa) 540 20CH3I + CO 650 20

" With no addition, the assay contained 7 mg of protein, 2.5 p.mol ofNADH, 25 U of LDH, 0.1 mM BES, and H2-CO2 gas.

the direction of methyl viologen reduction, was 4.3 ,umol ofmethyl viologen reduced min-1 mg of protein-1. CO dehy-drogenase was also a major protein in extracts of the closelyrelated bacterium Methanococcus vannielii (5). Likewise,the methyl viologen-dependent pyruvate oxidoreductase ac-tivity was 70 nmol of methyl viologen reduced min-' mg ofprotein-1. These levels of activity were comparable to thosepreviously reported for these enzymes in methanogens. It isnoteworthy that the oxidative capabilities of both of theseenzymes far exceeded the biosynthetic or physiologicalactivities, which were determined from the rate of lactatesynthesis. Thus, the level of lactate biosynthetic activity inthese extracts was generally 2 to 3 nmol of lactate synthe-sized min-1 mg of protein-' in the presence of BES. In theabsence of BES, the activity was somewhat lower (Table 8).After growth with acetate, extracts of Methanococcus ma-ripaludis contained lower levels of CO dehydrogenase, py-ruvate oxidoreductase, and lactate synthesis activity thandid autotrophically grown cells (Table 8). These resultswould be expected if the synthesis of the enzymes forautotrophic CO2 fixation were repressed during heterotro-phic growth.

Inhibition by cyanide. CO dehydrogenase is sensitive tolow concentrations of cyanide, and inhibition by cyanide isdiagnostic of pathways which use this enzyme in manyanaerobes (4, 6, 9, 10, 35, 41). A low concentration ofcyanide, 0.2 mM, abolished autotrophic growth of Metha-nococcus maripaludis, and acetate partially restored growth(data not shown). Likewise, the concentration of cyaniderequired for 50% inhibition (150) of methyl viologen-depen-dent CO dehydrogenase activity was 0.3 mM. Much higherconcentrations of cyanide were required to inhibit autotro-phic acetyl-CoA biosynthesis measured by the lactate trap(Fig. 3). In the presence of 20% C02, the 150 was near 20mM. Because this concentration of cyanide also inhibitedlactate synthesis in the presence of an acetyl-CoA-gener-ating system, inhibition may have been due to formation ofthe cyanohydrin of pyruvate or another indirect effect.

TABLE 8. Enzymes associated with autotrophy inMethanococcus maripaludis

Sp act (nmol min-' mgof protein-') after

Enzymatic activity growth"Without Withacetate acetate

Acetyl-CoA synthaseb 1.25 0.42CO dehydrogenase (EC 1.2.99.2)' 4,300 95Pyruvate oxidoreductase (EC 1.2.7.1)c 66 28

"The results are averages of at least three determinations.b Rate of lactate synthesis in the absence of BES.'Methyl viologen dependent.

J. BACTERIOL.

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

METHANOCOCCUS AUTOTROPHY 3077

20% C02

0 10 20 50

[KCN] (mM)FIG. 3. Inhibition of autotrophic acetyl-CoA synthesis and py-

ruvate synthase by cyanide. (A) Extracts (4 mg of protein) wereassayed for autotrophic acetyl-CoA synthesis and pyruvate syn-thase with an LDH trap for pyruvate. (B) Extracts (3 mg of protein)were assayed for pyruvate synthase alone in the presence of an

acetyl-CoA-generating system. The CO2 atmosphere of the assay isgiven in percent.

However, in the presence of 1 or 5% CO2, the 150 foracetyl-CoA biosynthesis was reduced to about 2 or 5 mM,respectively (Fig. 3). At these concentrations, cyanide didnot inhibit lactate synthesis in the presence of the acetyl-CoA-generating system. Therefore, inhibition may be attrib-uted to a direct effect on acetyl-CoA synthesis. The require-ment for high concentrations of cyanide for inhibition maybe due to protection by CO2 in the assay and disruption ofthe enzyme system upon preparation of the extracts.Cyanide also inhibited pyruvate synthesis at low concen-

trations of CO2. Thus, at 1% CO2 the I50 was 10 mM (Fig.3B). Higher concentrations of CO2 were protective. Like-wise, in the methyl viologen-dependent assay, the 50 for thepyruvate oxidoreductase was 2 mM (data not shown).

DISCUSSION

The evidence that the autotrophic acetyl-CoA biosyn-thetic pathway is operative in Methanococcus maripaludis isas follows. (i) Cell extracts prepared from autotrophicallygrown cells contain high levels of CO dehydrogenase activ-

ity. This activity is essential for autotrophic CO2 fixation viathis pathway. Moreover, CO dehydrogenase activity isgreatly reduced in extracts of heterotrophically grown cells.(ii) Lactate synthesis from CO2 is stimulated by formalde-hyde, methyl iodide, and carbon monoxide. These com-pounds are C-1 donors for autotrophic acetyl-CoA biosyn-thesis (21, 31, 46). (iii) The pattern of 14CO2 incorporationinto lactate is consistent with the proposal that lactate wasformed in part by autotrophic acetyl-CoA biosynthesis.Moreover, 14CH20 and 14CO specifically radiolabel the C-3and C-2 carbons of lactate, respectively. (iv) Autotrophicgrowth and lactate synthesis are inhibited by cyanide, whichis a specific inhibitor of CO dehydrogenase at low concen-trations (4, 35). These results further confirm the presence ofthe autotrophic acetyl-CoA biosynthetic pathway in Metha-nococcus maripaludis.However, a great deal of uncertainty remains concerning

the nature of this pathway in methanococci, as well as inother methanogens. For the most part, arguments for thepathway in methanogens depend heavily upon analogieswith the elegant work on C. thermoaceticum by Ljungdahland Wood et al. (32, 50). For instance, the role of a corrinoidprotein in methyl transfer has never been demonstrateddirectly in methanogens. Moreover, fundamental differencesbetween the methanogenic and clostridial system may exist,especially in the activation of CO2. In clostridia, CO2 isactivated by formate dehydrogenase and formyl-tetrahydro-folate synthetase. ATP is required for this latter enzyme(32). In methanogens, formyl-THMPT is synthesized with-out a stoichiometric requirement for ATP (M. I. Donnellyand R. S. Wolfe, Abstr. Annu. Meet. Am. Soc. Microbiol.1986, 1-56, p. 174). Instead, activation of CO2 appears to becoupled to the methylreductase system (19, 37). Therefore,the pathway of autotrophic acetyl-CoA biosynthesis, whichremoves intermediates of methanogenesis for cell carbonsynthesis, must also restore C-1 intermediates if methano-genesis is to continue. The nature of these anapleroticreactions is not known.

Lactate synthesis in these extracts underestimates the truerate of autotrophic CO2 assimilation by 1 or 2 orders ofmagnitude. Assuming a cellular protein content of 70% andthat 60% of the cell carbon is derived from acetate (40), therate of acetate synthesis necessary to support the specificgrowth rate of 0.27 h-1 during autotrophic growth is 80 nmolmin-' mg of protein-1. However, considering the complex-ity of these reactions and the sensitivity of these catalysts tominute quantities of oxygen, the low in vitro activity was notunexpected. Low rates of acetyl-CoA biosynthesis from H2and CO2 (0.1 to 0.2 nmol min-1 mg of protein-1) were alsoobserved in Triton X-100-treated cell suspensions, as well asin cell extracts of Methanobacterium thermoautotrophicum(44). Moreover, even in an ammonium sulfate fraction of anextract of Methanobacterium thermoautotrophicum, therates of acetyl-CoA biosynthesis were only 15 nmol min-'mg of extract protein-1 (31).Methanogens are a very diverse group of archaebacteria

(51). Previous studies of carbon fixation in methanogenshave focused on Methanobacterium thermoautotrophicumand Methanosarcina barkeri, which represent two of thethree major phylogenetic groups of methanogens. In thispaper, we report on carbon metabolism in Methanococcusmaripaludis, which represents the third phylogenetic groupof methanogens. We have established a simple spectropho-tometric assay to monitor the initial CO2 assimilation inextracts of Methanococcus maripaludis. By labeling andenzyme studies, we have demonstrated the presence of the

VOL. 170, 1988

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

3078 SHIEH AND WHITMAN

autotrophic acetyl-CoA pathway in this bacterium. There-fore, the acetyl-CoA biosynthetic pathway for autotrophicCO2 fixation is present in representatives of each majorphylogenetic group of methanogens.

ACKNOWLEDGMENTS

We thank J. C. Escalante-Semerena, L. G. Ljungdahl, C. C.Black, and N. L. Schauer for helpful discussions.

This work was supported by National Science Foundation grantDMB83-51355, Georgia Power Company, and the University ofGeorgia Research Foundation, Inc.

LITERATURE CITED1. Aronoff, S. 1956. Techniques of radiobiochemistry. The Iowa

State College Press, Ames.2. Bonner, W. M., and J. D. Stedman. 1978. Efficient fluorography

of 3H and "4C on thin layers. Anal. Biochem. 89:247-256.3. Bradford, M. 1976. A rapid and sensitive method for the

quantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal. Biochem. 72:248-254.

4. Daniels, L., G. Fuchs, R. K. Thauer, and J. G. Zeikus. 1977.Carbon monoxide oxidation by methanogenic bacteria. J. Bac-teriol. 132:118-126.

5. DeMoli, E., D. A. Grahame, J. M. Harnly, L. Tsai, and T. C.Stadtman. 1987. Purification and properties of carbon monoxidedehydrogenase from Methanococcus vannielii. J. Bacteriol.169:3916-3920.

6. Diekert, G., M. Hansch, and R. Conrad. 1984. Acetate synthesisfrom 2CO2 in acetogenic bacteria: is carbon monoxide anintermediate? Arch. Microbiol. 138:224-228.

7. Drake, H. L., S. I. Hu, and H. G. Wood. 1981. Purification offive components from Clostridium thermoaceticum which cata-lyze synthesis of acetate from pyruvate and methyltetrahydro-folate. J. Biol. Chem. 256:11137-11144.

8. Eden, G., and G. Fuchs. 1982. Total synthesis of acetyl coen-zyme A involved in autotrophic CO2 fixation in Acetobacteriumwoodii. Arch. Microbiol. 133:66-74.

9. Eikmanns, B., G. Fuchs, and R. K. Thauer. 1985. Formation ofcarbon monoxide from CO2 and H2 by Methanobacteriumthermoautotrophicum. Eur. J. Biochem. 146:149-154.

10. Eikmanns, B., and R. K. Thauer. 1984. Catalysis of an isotopicexchange between CO2 and the carboxyl group of acetate byMethanosarcina barkeri grown on acetate. Arch. Microbiol.138:365-370.

11. Escalante-Semerena, J. C., K. L. Rinehart, and R. S. Wolfe.1984. Tetrahydromethanopterin, a carbon carrier in methano-genesis. J. Biol. Chem. 259:9447-9455.

12. Escalante-Semerena, J. C., and R. S. Wolfe. 1985. Tetrahydro-methanopterin-dependent methanogenesis from nonphysiolo-gical C1 donors in Methanobacterium thermoautotrophicum. J.Bacteriol. 161:696-701.

13. Fuchs, G. 1986. CO2 fixation in acetogenic bacteria: variationson a theme. FEMS Microbiol. Rev. 39:181-213.

14. Fuchs, G., U. Schnitker, and R. K. Thauer. 1974. Carbonmonoxide oxidation by growing cultures of Clostridium pasteu-rianum. Eur. J. Biochem. 49:111-115.

15. Fuchs, G., and E. Stupperich. 1980. Acetyl CoA, a centralintermediate of autotrophic CO2 fixation in Methanobacteriumthermoautotrophicum. Arch. Microbiol. 127:267-272.

16. Fuchs, G., and E. Stupperich. 1982. Autotrophic CO2 fixationpathway in Methanobacterium thermoautotrophicum. Zen-tralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig.Reihe C 3:277-288.

17. Fuchs, G., and E. Stupperich. 1986. Carbon assimilation path-ways in archaebacteria. Syst. Appl. Microbiol. 7:364-369.

18. Fuchs, G., E. Stupperich, and R. K. Thauer. 1978. Acetateassimilation and the synthesis of alanine, aspartate and gluta-mate in Methanobacterium thermoautotrophicum. Arch. Mi-crobiol. 117:61-66.

19. Gunsalus, R. P., and R. S. Wolfe. 1977. Stimulation of CO2reduction to methane by methyl-coenzyme M in extracts ofMethanobacterium. Biochem. Biophys. Res. Commun. 76:790-

795.20. Gutmann, I., and A. W. Wahlefeld. 1974. L-(+)-lactate determi-

nation with lactate dehydrogenase and NAD, p. 1464-1468. InH. A. Bergermeyer (ed.), Methods of enzymatic analysis, 2nded. Weinheim, Deerfield Beach, Fla.

21. Holder, U., D. E. Schmidt, E. Stupperich, and G. Fuchs. 1985.Autotrophic synthesis of activated acetic acid from two CO2 inMethanobacterium thermoautotrophicum. III. Evidence forcommon one-carbon precursor pool and the role of corrinoid.Arch. Microbiol. 141:229-238.

22. Hu, S. I., H. L. Drake, and H. G. Wood. 1982. Synthesis ofacetyl coenzyme A from carbon monoxide, methyltetrahydro-folate, and coenzyme A by enzymes from Clostridium thermoa-ceticum. J. Bacteriol. 149:440-448.

23. Hu, S. I., H. L. Drake, and H. G. Wood. 1984. Acetate synthesisfrom carbon monoxide by Clostridium thermoaceticum. Purifi-cation of the corrinoid protein. J. Biol. Chem. 259:8892-8897.

24. Jansen, K., G. Fuchs, and R. K. Thauer. 1985. Autotrophic CO2fixation by Desulfovibrio baarsii: demonstration of enzymeactivities characteristic for the acetyl-CoA pathway. FEMSMicrobiol. Lett. 28:311-315.

25. Jansen, K., R. K. Thauer, F. Widdel, and G. Fuchs. 1984.Carbon assimilation pathways in sulfate-reducing bacteria. For-mate, carbon dioxide, carbon monoxide, and acetate assimila-tion by Desulfovibrio baarsii. Arch. Microbiol. 138:257-262.

26. Jones, W. J., D. P. Nagle, Jr., and W. B. Whitman. 1987.Methanogens and the diversity of archaebacteria. Microbiol.Rev. 51:135-177.

27. Jones, W. J., M. J. B. Paynter, and R. Gupta. 1983. Character-ization of Methanococcus maripaludis sp. nov., a new meth-anogen isolated from salt marsh sediment. Arch. Microbiol. 135:91-97.

28. Keltjens, J. T., G. C. Caerteling, C. Van Der Drift, and G. D.Vogels. 1986. Methanopterin and the intermediary steps ofmethanogenesis. Syst. Appl. Microbiol. 7:370-375.

29. Kenealy, W. R., and J. G. Zeikus. 1982. One-carbon metabolismin methanogens: evidence for synthesis of a two-carbon cellularintermediate and unification of catabolism and anabolism inMethanosarcina barkeri. J. Bacteriol. 151:932-941.

30. Lange, S., and G. Fuchs. 1985. Tetrahydromethanopterin, acoenzyme involved in autotrophic acetyl coenzyme A synthesisfrom 2CO2 in Methanobacterium. FEBS Lett. 181:303-307.

31. Lange, S., and G. Fuchs. 1987. Autotrophic synthesis of acti-vated acetic acid from CO2 in Methanobacterium thermoauto-trophicum. Synthesis from tetrahydromethanopterin-bound C,units and carbon monoxide. Eur. J. Biochem. 163:147-154.

32. Ljungdahl, L. G. 1986. The autotrophic pathway of acetatesynthesis in acetogenic bacteria. Annu. Rev. Microbiol. 40:415-450.

33. Ljungdahl, L. G., and H. G. Wood. 1982. Acetate synthesis, p.165-202. In D. Dolphin (ed.), Vitamin B12. Academic Press,Inc., New York.

34. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

35. Ragsdale, S. W., J. E. Clark, L. G. Ljungdahl, L. L. Lundie, andH. L. Drake. 1983. Properties of purified carbon monoxidedehydrogenase from Clostridium thermoaceticum, a nickel,iron-sulfur protein. J. Biol. Chem. 258:2364-2369.

36. Ragsdale, S. W., and H. G. Wood. 1985. Acetate biosynthesis byacetogenic bacteria. Evidence that carbon monoxide dehydro-genase is the condensing enzyme that catalyzes the final steps ofthe synthesis. J. Biol. Chem. 260:3970-3977.

37. Romesser, J. A., and R. S. Wolfe. 1982. Coupling of methylcoenzyme M reduction with carbon dioxide activation in ex-tracts of Methanobacterium thermoautotrophicum. J. Bacte-riol. 152:840-847.

38. Ruhlemann, M., K. Ziegler, E. Stupperich, and G. Fuchs. 1985.Detection of acetyl coenzyme A as an early CO2 assimilationintermediate in Methanobacterium. Arch. Microbiol. 141:399-406.

39. Sakami, W. 1955. Handbook of isotope tracer methods. West-em Reserve University, Cleveland, Ohio.

J. BACTERIOL.

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

METHANOCOCCUS AUTOTROPHY 3079

40. Shieh, J., and W. B. Whitman. 1987. Pathway of acetateassimilation in autotrophic and heterotrophic methanococci. J.Bacteriol. 169:5327-5329.

41. Smith, M. R., J. L. Lequerica, and M. R. Hart. 1985. Inhibitionof methanogenesis and carbon metabolism in Methanosarcinasp. by cyanide. J. Bacteriol. 162:67-71.

42. Stupperich, E., and G. Fuchs. 1981. Products of CO2 fixationand 14C labelling pattern of alanine in Methanobacterium ther-moautotrophicum pulse-labelled with 14Co2. Arch. Microbiol.130:294-300.

43. Stupperich, E., and G. Fuchs. 1983. Autotrophic acetyl coen-zyme A synthesis in vitro from two CO2 in Methanobacterium.FEBS Lett. 156:345-348.

44. Stupperich, E., and G. Fuchs. 1984. Autotrophic synthesis ofactivated acetic acid from two CO2 in Methanobacterium ther-moautotrophicum. I. Properties of in vitro system. Arch. Mi-crobiol. 139:8-13.

45. Stupperich, E., and G. Fuchs. 1984. Autotrophic synthesis ofactivated acetic acid from two CO2 in Methanobacterium ther-moautotrophicum. II. Evidence for different origins of acetatecarbon atoms. Arch. Microbiol. 139:14-20.

46. Stupperich, E., K. E. Hammel, G. Fuchs, and R. K. Thauer.1983. Carbon monoxide fixation into the carboxyl group ofacetyl coenzyme A during autotrophic growth of Methanobac-terium. FEBS Lett. 152:21-23.

47. Whitman, W. B. 1985. Methanogenic bacteria, p. 3-84. In C. R.Woese and R. S. Wolfe (ed.), The bacteria, vol. 8. AcademicPress, Inc., New York.

48. Whitman, W. B., J. Shieh, S. Sohn, D. S. Caras, and U.Premachandran. 1986. Isolation and characterization of 22mesophilic methanococci. Syst. Appl. Microbiol. 7:235-240.

49. Whitman, W. B., and R. S. Wolfe. 1983. Activation of themethylreductase system from Methanobacterium bryantii byATP. J. Bacteriol. 154:640-649.

50. Wood, H. G., S. W. Ragsdale, and E. Pezacka. 1986. Theacetyl-CoA pathway: a newly discovered pathway of autotro-phic growth. Trends Biochem. Sci. 11:14-18.

51. Woese, C. R. 1987. Bacterial evolution. Microbiol. Rev. 51:221-271.

52. Zeikus, J. G., G. Fuchs, W. Kenealy, and R. K. Thauer. 1977.Oxidoreductases involved in cell carbon synthesis of Methano-bacterium thermoautotrophicum. J. Bacteriol. 132:604-613.

VOL. 170, 1988

on May 7, 2021 by guest

http://jb.asm.org/

Dow

nloaded from